Determine if we want to introduce link indirection for Observer

Today Observer works predominately in the newer 5.5 kW power shelf. However, the PCBA itself is being designed such that it could backfill in the existing power shelf design. The main gotcha there is that the orientation there is different. Meaning that what we think of and describe externally as the two links to the outside world are labeled in an opposite fashion due to the differing orientation. There are numerous ways that we can address this and whether we actually use the optionality for the newer PSC is still a tbd. One thing we should consider and probably discard is whether or not we want to build a notion of what these links are with an indirection layer into hubris at the start of the Observer work. This likely doesn't make sense, but the idea here is that if the actual thing that we're plugged into can cause us to change the mechanical representation of what we're calling the various ix connectors, stats, and everything else we may want to be able to adjust all our reporting. We wouldn't want something to be labeled as say going to switch 0 and yet have it actually be something we call going to switch 1 internally.

There are plenty of other non-software ways to deal with this. One is not to change the labeling in the different orientation such that this is divergent from what used to be there but is electrically consistent. Mostly the ask is to make a conscious decision that we're going to probably not do this from the start.

No file, test, or entry point is named. First clarify whether Observer should support link indirection for differing power-shelf orientations, which connectors and reporting are in scope, and what decision or implementation outcome would count as done.